Jail-Breaking Parser Technology (A Slight Return)
2026-04-12
My view on parsers:
parsers are just string pattern matchers, i.e. the “easiest” parts of compilers
“regular expressions” used to be a compiler-only technology (e.g. LEX), then made the leap into other uses, appearing in places like UNIX tools (grep, sed, etc.), perl, javascript, python, etc. It is time for “parsing” to make that leap, too, but, 50 years after the release of YACC, this hasn’t really happened yet.
IMO, laborious CFG techniques discourage the thought processes for making this kind of leap whereas PEG techniques enable this kind of leap.
Making the leap into non-traditional compiler realms opens new doors and new views on programming techniques and even opens new doors on what we currently call “programming languages”.
IMO, current “programming languages” (GPLs - General Purpose Languages) are just variants of assembler. If we could simply stretch little syntactic skins over these assembler variants we could build new languages without needing to build more compilers from scratch, i.e. language design for the 21st century (IMO: SPL (Special Purpose Languages) not GPL, not-necessarily-Turing-complete, etc.) [ask me for my examples of little languages stretched over existing languages, if interested]
See Also
Email: ptcomputingsimplicity@gmail.com
Substack: paultarvydas.substack.com
Videos: https://www.youtube.com/@programmingsimplicity2980
Discord: https://discord.gg/65YZUh6Jpq
Leanpub: [WIP] https://leanpub.com/u/paul-tarvydas
Twitter: @paul_tarvydas
Bluesky: @paultarvydas.bsky.social
Mastodon: @paultarvydas
(earlier) Blog: guitarvydas.github.io
References: https://guitarvydas.github.io/2024/01/06/References.html
Paid subscriptions are a voluntary way to support this work.